Property Data Field Dictionary: What Each Column Means

A property export is one row per property and 46 columns. Most column names are self-explanatory until you need to make a decision with them: should a letter go to Address or Owner Mailing Address? Is Estimated Value what the county taxes the house on? Does a blank Open Lien Count mean there is no mortgage?

This dictionary goes through every column in a Bulk Property Records export, grouped the same way as on the home page, with what it means, where it comes from, and what to watch for when you use it. The records are compiled from county assessor and recorder filings across the U.S. and refreshed monthly.

The columns at a glance

The full set, by group. Column names are written exactly as they appear in the file’s header row: words with spaces, and a few with punctuation, such as Lot (Square Feet) and Vacant?. In SQL, put them in double quotes ("Owner Mailing Zip"); in pandas, use df["Owner Mailing Zip"], or rename them to short names once, right after loading.

# Property
Property ID,Address,City,State,Zip,County,APN,Property Type,Bedrooms,Bathrooms,Living Square Feet,Lot (Square Feet),Year Built,Units Count,Vacant?

# Owner
Owner 1 First Name,Owner 1 Last Name,Owner 2 First Name,Owner 2 Last Name,Owner Mailing Address,Owner Mailing City,Owner Mailing State,Owner Mailing Zip,Owner Type,Owner Occupied,Corporate Owned,Owner Property Count,Ownership Length (Months),Years Owned

# Value & equity
Estimated Value,Estimated Equity,Equity Percent,DTV,Free And Clear

# Mortgage
Mortgage Balance,Open Lien Count,Lien 1 Lender,Lien 1 Rate

# Sale
Last Sale Date,Last Sale Price

# Distress
Preforeclosure,Foreclosure Status,Foreclosure Auction Date,Tax Delinquent Year,Motivation Score

# File
Data Version

Two sources, two kinds of fact

It helps to know which public office a column ultimately comes from, because that tells you how it behaves.

• Assessor data describes the parcel for tax purposes: its identifier, its location and its physical characteristics. Assessors update on their own cycles, often once a year.
• Recorder data comes from documents filed with the county: deeds, mortgages, releases and foreclosure notices. It changes when someone records something, which can happen any day.
• Derived values are calculated from the other two: estimated value, equity, debt-to-value, the yes/no flags and the motivation score. They are useful shortcuts, but they are estimates, and they can change from one monthly refresh to the next even when nothing was recorded.

Property

APN and county: identifying a parcel

There is no national parcel number. Each county assigns its own APN, so the same APN can exist in two counties, and the same county name can exist in two states. The combination State + County + APN is what identifies a parcel as the county knows it. Within your own files, Property ID is the simpler key to use when you load the data into a database.

If you need a numeric county code, the file does not include one. Join State and County to the Census Bureau’s county reference file (national_county2020.txt on census.gov) to get the FIPS code. Its county names include the “County” or “Parish” suffix, so normalise the names on both sides before you join.

APN formats vary: some counties use plain digits, others dashes, dots or letters. Keep the value exactly as delivered. If you need to match against a county website that formats APNs differently, create a second, normalised column (for example, with punctuation stripped) instead of changing the original.

Owner

Property address vs mailing address

This is the distinction that matters most for outreach. Address, City, State and Zip are where the property is. The Owner Mailing columns are where the owner receives mail, taken from the address the county uses for the tax bill.

• For an owner who lives in the home, the two are usually the same.
• For a landlord, a second-home owner or a company, the mailing address is somewhere else: another house, an office, a PO box, or a property manager.
• If you are writing to the owner, use the mailing address. A letter sent to a rental property reaches the tenant.

Comparing the two addresses is also how absentee ownership is generally determined. The file has no separate absentee column; Owner Occupied = false is the closest approximation. It is derived, so when it matters, compare Address with Owner Mailing Address yourself and look at the records where your comparison and the column disagree.

Owner names

Owner names are recorded as the county has them and split into first and last name fields. Expect individuals in several formats, joint owners, and entities such as LLCs, trusts and estates, whose names do not split neatly into first and last. Corporate Owned and Owner Type mark owners that appear to be businesses. For mail merges, decide on a salutation strategy before you start; raw recorder names rarely read well in “Dear …” without cleanup.

Value and equity

Estimated value vs last sale price

A record carries two dollar values for the property, and they answer different questions:

• Estimated Value answers “what might this sell for now?” It is modelled from sales and property characteristics and is the input to Estimated Equity, Equity Percent and DTV.
• Last Sale Price answers “what did it last change hands for?” It is a recorded fact, but it can be many years old, and some transfers record no price or a nominal one.

Neither is the value the county taxes the property on. Assessed values are not part of the export.

Equity, Equity Percent and DTV

All three come from the same two inputs: the estimated value and the estimated balance of open liens (Mortgage Balance). Because of that, they move together. As the balance falls or the value rises, DTV goes down and Equity Percent goes up. If you want to verify the relationship in your own copy, recompute it:

-- After loading an export into a table named properties with the
-- header names kept as they are (DuckDB does this by default):
-- recompute equity from its inputs and compare with the delivered value.
SELECT "State", "County", "APN",
       "Estimated Value", "Mortgage Balance", "Estimated Equity",
       "Estimated Value" - coalesce("Mortgage Balance", 0) AS equity_check
FROM properties
WHERE "Estimated Value" IS NOT NULL
LIMIT 20;

Both inputs are estimates. Treat equity as a way to rank and segment records, not as a figure to quote to an owner.

Mortgage

What “open” means

A lien is open when a loan has been recorded and no matching release or satisfaction has been recorded. Liens drop off when the payoff is recorded. That depends on the county receiving and indexing the release, so a loan that was paid off recently can still show as open for a while.

Mortgage Balance is an estimate of what is still owed across all open liens, not what was borrowed and not a payoff figure. If you need a payoff figure, it has to come from the lender, not from public records.

Sale

Not every transfer is a sale

The last recorded transfer is not always an arm’s-length sale. A transfer into a living trust or between family members may record a price of zero, and it can be more recent than the purchase that set the owner’s real cost.

When you analyse prices, filter out zero and nominal values first, and treat a blank price as unknown rather than zero. Some states do not require sale prices to be disclosed on recorded deeds, so blanks cluster by state.

Distress

File

Yes/no columns and building lists

The true/false columns pre-compute segments people often build. They save time, but each one is a rule applied to the other columns or to recorded documents, so it is worth knowing roughly what each one means.

Two common segments have no column of their own. For absentee owners, filter on Owner Occupied = false, an approximation as described above. For high equity, pick your own cutoff on the underlying column, for example Equity Percent >= 50. A tax-delinquent list is the records with a non-blank Tax Delinquent Year.

Event-based columns such as Preforeclosure and Tax Delinquent Year reflect what had been recorded or reported when the monthly dataset was built. A situation can be resolved between refreshes, so treat these as reasons to look closer, not as statements about the owner today.

Blanks, zeros and units

• A blank means unknown. Not every county supplies every field. A blank Bedrooms is not zero bedrooms, and a blank Open Lien Count is not proof that a property is free and clear.
• Yes/no values are text. They are written as true and false; dates are YYYY-MM-DD.
• Money is in U.S. dollars. Values, balances and sale prices are whole dollars.
• Percentages are 0–100. Equity Percent and DTV of 40 mean 40%, not 0.4.
• Area is square feet. Both Living Square Feet and Lot (Square Feet).
• Every export is from a specific month. The Data Version column records which; it matters when you compare against a later export.

What the data is for

These records support business uses such as market research, investment analysis, prospecting and direct mail, subject to the laws that apply to your outreach. They may not be used for any purpose covered by the Fair Credit Reporting Act, such as tenant, employment, credit or insurance screening. Bulk Property Records is not a consumer reporting agency.

For a worked example that puts several of these columns together, see how to build a county-wide mailing list from a property export.